Personal electronic devices such as personal computers, personal digital assistants (PDAs), wireless telephones and pagers have become prevalent in recent years. These devices allow mobile computing by communication over wireless and/or wireline networks. The networks provide interconnection of these mobile devices with information sources as well as other similar devices. Commonly, the networks include communication over the Internet.
Typically, mobile computing devices include some form of user interface (UI). The nature of the user interface on a device is a function of the computing power and the hardware represented on the device. For example, a wireless telephone has limited computing power and a relatively small graphical user interface (GUI) with limited graphics and user interface capability. Conversely, a notebook personal computer has relatively extensive graphics capability, user interface capability and computing power. As the popularity of mobile devices increases, the main device platform for applications may shift from desktop personal computers to mobile devices.
One significant difficulty with developing applications for mobile devices is device heterogeneity. Due to variations in display size, display resolution, command input methods and GUI libraries, an application developer may have to re-design and re-implement applications for the graphical user interfaces (GUIs) of each device platform. With the large number of different mobile devices currently available, or coming into the market, re-design and re-implementation may be an ongoing, labor intensive and cost prohibitive endeavor.
One solution is through the development of model-based techniques. An exemplary model-based technique is user interface modeling. In general, user interface modeling involves a platform model, a presentation model and a task model. The platform model describes the operational functionality forming the user interface for each device supported. The presentation model, describes the hierarchy, stylistic choices, selection and placement regarding the visual appearance of the user interface associated with the supported device(s). The task model identifies tasks that a user of the supported device(s) may perform. With this technique, mappings between the various models may be developed to produce a user interface for a particular device.
Implementation of the model-based approach typically includes development of a high level language to implement the models. In addition, developers typically build and specify significant portions of the models to support the devices. Since the high level language is fairly complex, developers usually must learn the language, as well as the execution mechanics, prior to implementing the model-based approach. Further, model-based approaches typically generate code based on the different model(s) that the developer builds. Slight differences in the user interface requirements of a device may result in considerable differences in the code generated for two seemingly similar devices. Accordingly, the complexity and level of programming expertise required by the developer is significant.